Research Design

In this paper, we aim to examine how corporate governance has changed within the banking sector following the course of events that took place during and following the financial crisis, and how the events were affected by institutional change. The survey could take many different forms where qualitative and quantitative methods each have their own merits. Yet, given our open-ended question, we avoid the use of empirical surveys that are structured around hypothesis formulation and statistical falsification tests. Rather, we choose on a more qualitative assessment where the methodology is tied to our fundamental assumptions stated in previous section.

Figure 4

The comments below explain our analytical approach and the connections between a bank’s governance and its surroundings 1. Identify corporate governance categories

1. We use agency theory to arrive at a set of key facets of a bank’s overall corporate governance structure

2. To make the study operational, we construct variables which work as proxies for the characteristics and practices of interests 3. The variables are arranged into four groups in order to facilitate the understanding of, and to add structure to, our analysis 4. For each of the variables we map the state in two periods, (1) pre crisis period 2003 – 2007 (2) post crisis period 2008 – 2011.

Data is collected and compared at three observation points, 2003, 2007 and 2011, enabling to distinguish potential trend shift in the variables driven by the financial crisis

2. Identify stakeholder classes and subsequently how stakeholder pressure exerted on banks changed following the crisis

1. Based on stakeholder theory, we categorize banks’ stakeholders into seven groups based upon their individual characteristics 2. Changes which stakeholder groups undergo due to the financial crisis and which have an impact on its relationship to the banks

is analysed

3. Pressures from stakeholders translate into change in one or several of the three institutional pillars

1. Depending on the identity of the stakeholder and the means through which the pressure is exercised, stakeholder influences can affect three overarching institutional pillars; regulative, normative and cultural-cognitive

2. In the interaction between banks and its stakeholder environment, institutional change pushes banks to adapt their governance practices

4. Corporate governance practices are affected by changed pressures from stakeholders after the outbreak of the financial crisis 1. Subject to the pressures exercised by its stakeholders following the financial crisis, we map and analyse changes to banks’

corporate governance practices

2. The connection between changes observed in corporate governance practices following the crisis and institutional change is further strengthened by the use of case studies

2

2003

2011 2007

Before FinancialCrisis D. InternalAgency Structures C. Board Compositionand Committees B. ExecutiveCompensation A. OwnershipStructure

Stakeholders

1. Dormant

2. Discretionary

3. Demanding

4. Dominant

5-Dangerouns

6. Dependent

7. Definitive

Corporate Governance

D’ C’ B’ A’ 1’2’3’4’5’6’7’

Regulative Normative Cultural-Cognitive

Post Crisis Outbreak

Institutional Change 2.

1.

3.

4.

30 Figure 4 above outlines the key traits of our analytical methodology. The following sections will elaborate on our methodological approach and its motivation.

3.1.1 Determinants of methodological choice

Indeed, any research methodology must be consistent with the underlying theoretical orientation of the researcher, and with the research question at issue. The methodology is also linked to the nature of the phenomenon to be studied, and there are broad currents that have been acclaimed as the “correct”

mode of conducting research within a particular area. In addition, the insight sought dictates the methods through the way it can be obtained. As such, operating in an area that is traditionally dominated by functionalist researchers, but with a research question that proposes a context-related problem, calls for a conscious choice of methodology. Migrating from a purely functionalistic field towards an interpretive world-view, we select a relatively novel and unexplored, yet resourceful moment to our paper: the clinical approach to research in finance (Tufano 2001).

3.1.2 Clinical opposed to scientific research

While a formal definition of clinical research is still to be agreed upon, the term commonly refers to empirical work in which a relatively small set of observations is used. Introspection tends to be of minor importance, and insights are driven by observation. Data can be either public or private, but the shared characteristic of all clinical analyses is intensive analysis of the results and efforts towards disaggregation. The approach is tailored to research founded upon a small number of cases of particular interest although its applicability is considerably broader today (Jensen et al, 1989, p. 3). In contrast, traditional scientific research within the functionalist paradigm encompasses methodologies where logical propositions are developed through introspection or mathematical theory. The theorems are translated into hypotheses which are tested using large datasets, and implicit in the methodology is the researchers confidence in the set of variables on which she chooses to focus (Tufano 2001, pp.

181-3).

Instead of contrasting the two branches, as does e.g. Ardalan (2007), we view clinical methodologies as a first stage in functionalist, scientific research rather than a research methodology in itself. This explorative approach is intended to understand organizations beyond aggregation and encompasses a broad range of techniques. Its advantages over traditional research approaches are to be found in the development of new theory where the area of study covers human- or firm behaviour. This process, commonly referred to as adduction has i.a. been applied within the field of behavioural finance, and primarily through direct observation (Tufano 2001). The interplay between high-level empirical research and small-scale analysis that considers the fragmental mechanisms that underlie the aggregate associations has been brought forward as an important catalyst for the development of research and we view clinical research within the scientific sphere as a prime example of this (e.g. Karen Wruck in Ardalan 2003, p. 1046).

31 3.1.3 Motivating of choice of methodology

The intended contribution of our work is to reach beyond traditional aggregate associations. Pure scientific research can be rightfully applied only in situations where the researcher is confident in deducing reality through the use of predetermined variables. As we approach the situation with few preconditions, while we also believe that contextual aspects may influence the development of the corporate governance environment, we need to apply aspects from the branch of clinical methods.

To understand the mechanisms better, and to document the drivers of change, it is instructive to contextualise governance decisions and zoom in on real decisions made. We propose that a research approach where the micro events are analysed is necessary to drive understanding of phenomena in the aggregate. A key principle in our functionalist orientation is that all aggregate trends originate from individual acts, and we therefore note that the fundamental drivers behind governance decisions need to be mapped before any generalizations and structured tests can be made. An analysis of cause and effect in the decision making process is not only informative, but necessary to develop theory. Our approach allows us to contest prescriptive theories with actual developments and to comment on discrepancies in an unambiguous (albeit case-specific) fashion (Tufano’s opening speech at the conference, Ardalan 2003, p. 1045; Karen Wruck in Ardalan 2003, p. 1046). Further, since the very core of our research topic stems from human decision making, we hope to achieve the additional insights that behavioural finance scholars have successfully done in their use of clinical surveys. This is why we embrace the clinical method as a complement to the scientific method. Rather than providing a substitute for traditional scientific research methods, we complement and enhance the traditional methodology by introducing case based observations in our thesis.

Further, the relevance of clinical research is evermore present in periods of radical change in the conduct and organization of general institutions where roles and activities of economic actors change radically and traditional theory may fail. In such situations, clinical research steps in to evolve finance through observations of actual events. Our insights, where generalizable, can hopefully constitute a starting point for more sophisticated hypothesis formulation going forward within the field of corporate governance in periods of turmoil. In particular, when the environmental setting sees periods of turmoil, it may be the case that fundamental questions need to be reformulated, or that some of the explanations derived from observing a sequence of actions from a distance may no longer apply. As such, we will comment on the adequacy of current theories within the field of corporate finance for explaining the observed behaviour and determine the extent to which the current models are reasonable representation of the actual business settings. This is similar to what Krigman, Shaw &

Womack (2001) did when they examined the validity of large sample inference. Using surveys, the authors fortified their insights by adding primary information interpretations to the general associations. The article suggests the advantages of pairing different methodologies when looking at a single question which we hope to benefit from in our analysis.

32 In short, we infer that a scientific approach with clinical research applied to the explanatory discussion is the most suitable methodology for our paper given the importance of institutional arrangements for governance structures.

3.1.4 Our use of theory and cases

Our analysis originates from theory in the sense that we use previous insights to arrive at a set of guiding predictions and indications of what parameters should be the focus of our survey. Also, we use theory to structure the discussion of the institutional change that has taken place over the time period of interest. This theoretical account is not intended as a firm framework in a delineating sense, but rather as initial guidance and a starting point for sorting and filtering within the dataset.

Given the qualities of clinical research within the explanatory stage of scientific research, we hope to generate a more thorough and holistic understanding of the actual shift in the governance regime by bringing cases into the analysis. In particular, we will attempt to identify general trends, but also seek explanations for the most significant and surprising deviations within our selected dataset. Thereby, we try to overcome what Jensen et al (1989, p. 4) refer to as “possibility theorems” – propositions that are logically consistent but lack the ability to explain any real-world phenomenon.